Part of the Site Migrations series. For the full picture, start with Site migration SEO checklist: crawl before, during and after launch.
A few days after launch, the traffic line dips. The question that matters is whether it’s the normal turbulence of Google re-learning your site, or a mistake that will keep costing traffic until someone finds it. The two look similar for a week or two, and then they don’t.
This guide shows how to tell them apart and how to find the usual causes fast. For drops unrelated to a migration, see why organic traffic drops.
Is the drop normal?
Google’s guide to site moves sets expectations: rankings can fluctuate while Google recrawls and reindexes, and for medium-sized sites it can take “a few weeks or more for Google to gradually start showing the new URLs instead of the old ones.”
| Pattern | Likely meaning |
|---|---|
| Shallow dip across the site, recovering week by week | Normal reprocessing |
| New URLs gaining impressions as old ones lose them | Normal transfer in progress |
| Deep drop in specific sections or templates | Something broke there |
| Old URLs losing impressions, new URLs not gaining | Redirects or indexability problem |
| No recovery after several weeks | A cause that won’t fix itself |
Step 1: See where the traffic went
In Search Console’s Performance report, compare the weeks after launch with the weeks before, grouped by page. If the domain changed, look at both properties. Sort by click difference.
- Old URLs that lost clicks: where do they redirect now, and does the destination rank?
- New URLs: are they gaining impressions? If not, are they indexed?
- Sections: is the loss concentrated in one template or directory?
Step 2: Test every old URL
This is where most post-migration problems are. Crawl the list of old URLs (from your pre-launch benchmark crawl, or exported from Search Console) in Crawlens’s URL list mode:
| Check | What it means |
|---|---|
| Internal pages return 4xx | Old URLs with no redirect: their traffic is gone |
| Redirect chains (2+ hops) | New rules stacked on old ones |
| Redirect loops | URLs that never resolve |
| Temporary redirects | 302/307 instead of a permanent redirect |
| Pages with search traffic return errors | The valuable URLs that broke, with their clicks |
Then look at where the redirects go: export the crawl and check the Redirects to column against your redirect map. Many old URLs landing on the homepage or a broad category is a red flag; Google warns that can “be treated as a soft 404 error.” See redirect mapping.
Step 3: Check that the new pages can rank
Crawl the new site from its homepage. The mistakes that make new pages invisible:
- A staging
noindexor robots.txt rule left on: Noindex pages, Linked pages blocked by robots.txt. - Canonicals pointing at old or staging URLs: Canonicalised pages, Canonical points to a non-200 URL.
- Pages with traffic that aren’t indexable: Pages with search traffic are not indexable.
- Internal links still pointing at old URLs: Internal links point to redirects.
- Sitemaps listing old or redirected URLs: Non-indexable URLs in sitemaps.
Use URL Inspection on a few important new URLs to see whether Google has indexed them and which canonical it chose. In Crawlens, Inspect URL on a URL’s Search tab does the same from inside the crawl.
Step 4: Compare what the pages say
If redirects and indexing are fine, the content may have changed. Compare the new site with the pre-launch benchmark crawl. If URLs didn’t change, Crawlens’s Compare lists pages whose title changed, that became non-indexable or changed status, plus new issues. If URLs changed, compare matching templates by hand: titles, H1s, word counts, internal links, structured data. Lost text blocks, missing structured data and thinner templates explain drops that redirects don’t. See CMS migration SEO.
Also check rendering if the new front end is JavaScript-heavy (Content depends on JavaScript) and performance (Core Web Vitals).
Step 5: Fix in order of lost traffic
- Redirects for URLs with clicks that 404 or go to the wrong place.
- Indexability blockers: noindex, robots.txt, wrong canonicals.
- Chains and temporary redirects: collapse to one permanent hop.
- Internal links and sitemaps: point them at new URLs.
- Content and template gaps: restore what was lost.
Request indexing for the most important fixed URLs and resubmit the sitemap. (The Indexing API is not a shortcut here: it only covers job postings and livestreams.)
Step 6: Confirm the recovery
Re-crawl the old URL list and the new site to confirm the fixes are live. Schedule crawls with alerts for the next weeks, so a later deployment doesn’t undo them. In Search Console, watch new URLs gain impressions week by week. Keep the redirects for at least a year, as Google recommends.
For the full process to follow on the next migration, see the site migration SEO checklist.
Checklist
- Drop compared with Google’s expectation for normal fluctuation
- Search Console compared before and after launch, by page (both properties if the domain changed)
- Old URL list crawled: no 4xx, chains, loops or temporary redirects
- Redirect destinations checked against the map; no mass redirects to the homepage
- New site crawled: no staging noindex, robots.txt blocks or wrong canonicals
- Internal links and sitemaps use new URLs
- Templates compared with the benchmark: titles, content, links, structured data
- Fixes made in order of lost traffic
- Re-crawl confirms fixes; scheduled crawls watch for regressions
Frequently asked questions
Is it normal to lose traffic after a site migration?
Some fluctuation is normal. Google says rankings can fluctuate while it recrawls and reindexes, and that for medium-sized sites it can take a few weeks or more for the new URLs to replace the old ones. A drop that's deep, lasts longer or affects only some sections usually has a fixable cause.
How long does it take to recover after a migration?
Google says that for medium-sized sites it can take a few weeks or more for the new URLs to replace the old ones; larger sites take longer. Recovery only starts once the cause is fixed, so finding broken redirects or noindexed templates quickly matters more than waiting.
What are the most common causes of traffic loss after a migration?
Old URLs that return 404 or redirect to the homepage, redirect chains, a noindex or robots.txt rule left over from staging, canonicals pointing at old or staging URLs, and content, titles or internal links lost in new templates.
Should I roll back a migration if traffic drops?
Rarely. Rolling back is a second migration with its own risks. Find and fix the specific problems first; most can be fixed within days. Consider rolling back only if the new site is fundamentally broken and can't be fixed quickly.
How do I check which old URLs lost traffic?
In Search Console, compare the weeks before and after launch, grouped by page, and sort by the click difference. Then check where each of those old URLs now redirects, and whether the destination is a working, indexable equivalent.